Examination of Gender-Related Differential Item Functioning Through Poly-BW Indices.

Tsai-Wei Huang,Magdalena Mo Ching Mok,Pei-Chen Wu

doi:10.3389/fpsyg.2022.821459

Abstract

The existing differential item functioning (DIF) detection approaches relying on item difficulty or item discrimination are limited for understanding the associates of DIF items, and consequently, DIF items were conventionally either deleted or ignored. Given the importance of minimizing DIF items in test construction, teachers or testing practitioners need more information regarding possible associates of DIF items. Using an example of a teacher-made mathematics achievement test, this study aimed to examine how the Poly-BW indices (power, defenselessness, disturbance, and hint) contributed to the properties of gender-related DIF items. Data from a 34-item mathematics achievement test that involved 1,439 seventh-grade students from Taiwan (51.01% boys and 48.99% girls) showed that the differences of the defenselessness (mp) and power (cp) indices between men and women served as salient predictors of the DIF measures estimated by the Poly Simultaneous Item Bias Test (Poly-SIBTEST) procedure and with satisfactory accuracy of hit rates. Items with relatively large defenselessness for men were likely to present male-favoring DIFs, whereas items with relatively large power for men were likely to present female-favoring DIFs. The Poly-BW indices yielded directions for modifying items for teachers in practice.

Highlights

Differential item functioning (DIF) indicates the situation where participants from different memberships on the same level of the latent trait have a different probability of a certain response to a particular item (Holland and Thayer, 1988)
Two Poly-BW indices explain almost 78% of the variance (63.8% from the predictor of mpM−F and 14.2% from the predictor of cpM−F) of the differential item functioning (DIF) measures (Bu) obtained from PolySIBTEST analysis (F4,29 = 28.53, p < 0.001)
The results showed that three Poly-BW indices explain almost 88% of the variance (70% from the predictor of mpM−F, 15% from the predictor of cpM−F, and 3% from the predictor of wM−F) of the DIF contrast obtained from Rasch DIF analysis

Summary

Introduction

Differential item functioning (DIF) indicates the situation where participants from different memberships (e.g., age and gender) on the same level of the latent trait (e.g., math performance) have a different probability of a certain response to a particular item (Holland and Thayer, 1988). Most DIF detection approaches and the estimation of DIF effect sizes are limited in providing the possible associates of DIF items (Kim et al, 2007; Li, 2014a,b, 2015). The possible associates of DIF-flagged items entail the possible. Gender-Related DIF Through Poly-BW Indices reasons why the DIF occurs and through which we could determine whether these items should be deleted or revised. If the information on possible associates of DIF items is limited, the DIF items cannot be treated appropriately. Some researchers chose to remove a DIF-flagged item from the item bank, while others might conduct further analyses (e.g., Wang et al, 2012; Chen and Hwu, 2018). By reviewing 27 studies on DIF item treatment, Cho et al (2016) found that 30% of the studies removed the DIF items while 26% of the studies ignored them

Objectives

Methods

Results

Conclusion